We present a polarization-entangled photon pair source operating in the visible light range around 532 nm. Employing a collinear crossed-crystal scheme with type-I degenerate phase matching in barium borate (BBO), our source achieves a brightness of 9.5 k pairs/s/mW and a quantum state fidelity of 98.3%, making it a candidate for integration in microscopes and make use of the advantages of mid-visible optimized single-photon detection technologies. In order to study potential applications, we present a trade-off between source brightness and polarization entanglement visibility and propose use cases for different filtering configurations of the source, capable of a brightness up to 1.23 M pairs/s/mW.
Entangled photons are a key resource in photonic quantum technology, acting as low-noise probes in imaging and sensing, as versatile information carriers in information processing and communication networks or as tamper-proof padlocks for cryptography. Reliable and robust entangled photon sources (EPS) are, thus, an essential building block toward a manifold of applications.1 Among the many technologies that have been demonstrated, the most established mechanism for generating photon pairs with high state-fidelity in polarization, spatial modes, or temporal modes is through spontaneous parametric downconversion (SPDC).2 The majority of sources reported to date generate photons in the NIR to telecom wavelength range.3–8 Sources of entangled photons operating at shorter wavelengths could be advantageous for long-distance free-space transmission due to lower diffraction loss9 for quantum metrology and sensing with short-wavelength illumination.10–13 Moreover, for a practical implementation in terms of quantum microscopy, a close match to microscopes' optical design wavelength of 550 nm is of great advantage.
Recently, EPS at 532 nm have been reported,14–17 but the maximum brightness achieved was 135 pairs/s/mW,16 a value that we increase by almost two orders of magnitude thanks to the use of a more efficient collinear EPS scheme and an optimization SPDC focus and collection modes. It is noteworthy that in the literature, other time-correlated pair sources with wavelengths centered around 532 nm have been used due to the proximity to the peak of the human eye detection capability at 505 nm to prove the sensitivity of the human eye to single photons.18–20 Due to the polarization entanglement of our presented source, in principle, it could be used for recently proposed experiments, giving a human observer a direct role in detecting the predictions of standard quantum mechanics and even testing local realism.21
Another important benefit of short wavelength operation is the widespread availability of efficient single photon avalanche diodes (SPADs) with very high timing resolution at wavelengths near 550 nm. These SPADs with high efficiency , but more importantly, with timing jitter as low as 35 ps,22,23 provide a jitter much lower than the typical timing jitter between 300 and 1500 ps for wavelengths > 650 nm.24 Both the detection efficiency and the timing jitter are of great importance toward practical applications: on the one hand, detection efficiency leads to a higher generated key rate and faster image acquisition.25 On the other hand, timing resolution allows for smaller coincidence detection windows26 and, thus, decreases a number of false pair detections (accidental coincidences), which, in turn, leads to better visibility and correspondingly lower quantum bit error rates as well as improved signal-to-noise ratio in imaging.
Here, we report on the development of a short wavelength source of polarization entangled photon pairs with high brightness, which—combined with commercially available SPAD detectors with high timing resolution and efficiency—may pave the way to more cost-efficient systems for quantum communication, imaging, and sensing.
SPDC in bulk crystals offers a wide range of experimental arrangements to generate polarization entanglement.2 We aim at designing an EPS ready to be field compatible in future iterations, and for this reason, we choose a phase-stable EPS design, in which a common path configuration for down-converted and pump photons ensures that changes in the optical path length are experienced by the pump laser as well as the down-converted pair—leading to phase stability. Thereby two very popular approaches are the Sagnac loop,5 where a single down-converter is bi-directionally pumped inside a polarization interferometer, and the crossed-crystal scheme,27 in which two parametric down-converters, rotated by 90 with respect to each other, are placed in sequence and pumped with a diagonally polarized pump laser. Another practical consideration is that the Sagnac and other polarization-entangled photon schemes require the use of dual-wavelength components such as waveplates or polarizing beam splitters.4,5,7 Due to the limited spectral performance of dual-wavelength components for the short-wavelength pump and SPDC photons, and given the relatively broad SPDC spectrum for type-I phase-matching, we employ the collinear crossed-crystal scheme27 that does not require any custom multi-wavelength components. Recently, crossed-crystal schemes that offer a good overlap of horizontal and vertical polarization modes have been used to generate entangled photons at 532 nm.15,17 In our case, we also use the collinear case of SPDC emission, which has the advantage with respect to SPDC cone emission that a large portion of the beam can be efficiently coupled to single-mode fibers.
The setup scheme can be seen in Fig. 1, and the crystals used were two 1.5 mm BBO crossed crystals for SPDC generation cut at 47.66, two 0.75 mm BBO crossed crystals cut at the same angle for spatial walk-off compensation, and an yttrium vanadate crystal of 0.63 mm for temporal compensation, in the same orientations as in Ref. 27 In our setup, both signal and idler photons were collected in the same single mode fiber, so that it could be directly integrated in a quantum-enhanced microscopy scheme.10,28
Scheme of the crossed crystal source used in the lab. HWP: 266 nm half-wave place; L1 = 200 mm focusing lens; L2 = 75 mm collection lens; YVO4: yttrium vanadate temporal compensation crystal; 532 nm BPF: 24 nm FWHM bandpass filter. The source used two pairs of BBO crystals: two SPDC production crystals in the focus of the pump beam and two spatial compensation crystals to improve the SPDC beam overlap.27
Scheme of the crossed crystal source used in the lab. HWP: 266 nm half-wave place; L1 = 200 mm focusing lens; L2 = 75 mm collection lens; YVO4: yttrium vanadate temporal compensation crystal; 532 nm BPF: 24 nm FWHM bandpass filter. The source used two pairs of BBO crystals: two SPDC production crystals in the focus of the pump beam and two spatial compensation crystals to improve the SPDC beam overlap.27
In order to analyze the state produced by the EPS, the pairs were collimated again and split probabilistically by a beam splitter (BS) and focused into free-space single photon avalanche diodes for detection. By splitting the pairs probabilistically, it can be easily checked that only in half of the cases the signal and idler photons exit through different BS ports, and this results in a factor 1/2 of reduction of the measured pairs. Counting the possibilities to detect a single photon, it can be shown that using a BS reduces the number of detected single counts by a factor of 3/4. The brightness is given by the number of coincidences detected, and the heralding is as usual defined as the coincidences to single ratios, but in both cases, we add the inverse factors to compensate the probabilistic detection scheme. In order to characterize the entanglement, we place polarizers in front of the single photon detectors and measure coincidences CAB for polarizers orientations A, B in the horizontal and diagonal basis. We measure the visibilities for the H–V basis and for the D–A basis, and from there, we extract the bounded fidelity from .29
Naturally, the 532 nm EPS is pumped at 266 nm (cw), which causes a major challenge. At this spectral range, the very efficient periodically poled crystals, such as periodically poled potassium titanyl phosphate (ppKTP) and periodically poled lithium niobate (ppLN), are not transparent.30 As expedient, we use barium borate (BBO), which has an increased downconversion efficiency at such short pump wavelengths. Due to the walk-off present in type-I critical phase matching, the mode overlap approach to calculate lens parameters yielding a high brightness EPS is not easily calculated as in quasi-phase matched crystals.31,32 In order to increase the brightness, a number of focusing and collection lenses were experimentally tested to determine the most efficient combination for single mode fiber coupling, as seen in Fig. 2. The results are given in terms of the calculated Gaussian beams characterized by the Gaussian waist radii w0 in the center of the crystals.
Comparison of different spot sizes of the pump and collection mode to optimize EPS brightness, using a BBO pair of crystals of 1.5 mm length each. This comparison led to the decision of which focusing length to use for the crystal focusing and collection lenses.
Comparison of different spot sizes of the pump and collection mode to optimize EPS brightness, using a BBO pair of crystals of 1.5 mm length each. This comparison led to the decision of which focusing length to use for the crystal focusing and collection lenses.
After the experimental single-mode brightness optimization, the coupling was still greatly reduced compared to the total SPDC produced collected without single mode filtering. This shows that BBO crystals cannot generate rates comparable to ppKTP when coupling into the single-mode fiber, without other developments to better compensate for the walk-off effect.32 Nevertheless, pair coupling into multimode fibers showed much higher rates at the sacrifice of entanglement visibility. In order to demonstrate the high time-correlated pair generation, we adapted the source to the “correlated pair source” for photon pair generation only and used two 3 mm BBO crystals oriented in parallel so that photon pairs of only one polarization were produced but brightness was enhanced. The signal and idler were separated by a dichroic mirror and coupled into multi-mode fibers, and no spectral filtering was performed. This configuration would be useful for protocols, where quantum enhancement takes advantage of the correlations in time or spatial degrees of freedom such as enhanced two photon absorption,33 quantum lidar,34 or biphoton centroid microscopy.13
In order to study the trade-off between the brightness and entanglement visibility, we tested the polarization-entangled crossed-crystal in three filtering configurations: with a variable iris just after the collection lens and using free space detectors without any fiber coupling, and with single mode fiber coupling with and without spectral filtering. In the case of spectral filtering, a bandpass filter centered at 532 nm with FWHM = 24 nm wide enough to maintain most of the SPDC spectrum of about FWHM = 45 nm was used.
Table I shows a summary of how filtering configurations provide a trade-off between the polarization state fidelity and pair brightness.
Comparison of different degrees of spatial and spectral filtering. In the values with an asterisk, the degenerate case pair rate was measured by separating the signal and idler with a normal beam splitter and corrected with a factor of 2, and the heralding was corrected with a 4/3 factor. The average heralding is the mean ratio of coincidences to singles for the signal and idler channel. The correlated pair source uses parallel crystals to achieve time-correlated events, while the other ones use the crossed-crystal scheme.
Source . | Brightness . | Average heralding . | Fidelity . |
---|---|---|---|
Correlated-pair source | 1.23 M pairs/s/mW | 14.8% | ⋯ |
Iris filtering and spectral filtering | 132 k pairs/s/mW | 9.5% | 87.15% |
Single mode filtering | 30 k pairs/s/mW | 18.9% | 92.5% |
Single mode and spectral filtering | 9.5 k pairs/s/mW | 16.7% | 98.3% |
Source . | Brightness . | Average heralding . | Fidelity . |
---|---|---|---|
Correlated-pair source | 1.23 M pairs/s/mW | 14.8% | ⋯ |
Iris filtering and spectral filtering | 132 k pairs/s/mW | 9.5% | 87.15% |
Single mode filtering | 30 k pairs/s/mW | 18.9% | 92.5% |
Single mode and spectral filtering | 9.5 k pairs/s/mW | 16.7% | 98.3% |
In summary, we improved by almost two orders of magnitude the brightness for polarization entangled photon pairs at 532 nm using a collinear EPS scheme with optimized focusing and collection modes and studied how the polarization entanglement fidelity varies with different degrees of filtering. The source is used with cost-efficient low-jitter SPAD detectors that are optimized for the mid-visible (green) range and thereby benefits from current detection technology and enriches the application potential of photonic quantum technologies, in particular, within the field of quantum communication and quantum imaging. For example, in a recent proposal for an uplink to a 3U CubeSat, the key rate could be increased by over an order of magnitude using a source operating in the visible wavelength range.26
The authors thank Dr. Álavaro Cuevas of the Institute of Photonic Sciences (ICFO) for helpful discussions and Krishna Dovzhik of the Institute for Quantum Optics and Quantum Information (IQOQI) for preliminary tests on the use of different crystals with a 266 nm pump laser. This work has received funding from the European Union's Horizon 2020 FET-Open Research and Innovation Programme under Grant Agreement No. 801060 (Q-MIC project). Furthermore, this work was supported by a Fraunhofer LIGHTHOUSE PROJECT (QUILT) and by the Fraunhofer Attract program (QCtech).
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.